ITPEC FE Morning April 2020 Question 4

Supervised Learning — train on labeled data, predict on new inputs.

Supervised learning uses a dataset with known correct answers (labels) to train a model, which then predicts outputs for new, unseen input data.

• •Supervised — learns from labeled data (input → known output)
• •Unsupervised — finds hidden patterns without labels (clustering, association)
• •Reinforcement — learns by trial and reward (agent evaluates actions)

Why not others:
- (a) — describes unsupervised learning (no pre-existing labels)
- (b) — describes reinforcement learning (evaluating action results)
- (d) — describes behavioral prediction/analytics, not a learning paradigm definition

Key rule: "Trained on a dataset with answer keys" = supervised learning.

AI-generated — may contain errors